.beta.-Adrenoceptors have been subclassified as .beta..sub.1 and .beta..sub.2 since 1967. Increased heart rate is the primary consequence of .beta..sub.1 -receptor stimulation, while bronchodilation and smooth muscle relaxation typically result from .beta..sub.2 stimulation. Adipocyte lipolysis was initially thought to be solely a .beta..sub.1 -mediated process. However, more recent results indicate that the receptor-mediating lipolysis is atypical in nature. These atypical receptors, later called .beta..sub.3 -adrenoceptors, are found on the cell surface of both white and brown adipocytes where their stimulation promotes both lipolysis (breakdown of fat) and energy expenditure.
Early developments in this area produced compounds with greater agonist activity for the stimulation of lipolysis (.beta..sub.3 activity) than for stimulation of atrial rate (.beta..sub.1) and tracheal relaxation (.beta..sub.2). These early developments disclosed in Ainsworth et al., U.S. Pat. Nos. 4,478,849 and 4,396,627, were derivatives of phenylethanolamines.
Such selectivity for .beta..sub.3 -adrenoceptors could make compounds of this type potentially useful as antiobesity agents. In addition, these compounds have been reported to show antihyperglycemic effects in animal models of non-insulin-dependent diabetes mellitus.
A major drawback in treatment of chronic diseases with .beta..sub.3 agonists is the potential for stimulation of other .beta.-receptors and subsequent side effects. The most likely of these include muscle tremor (.beta..sub.2) and increased heart rate (.beta..sub.1). Although these phenylethanolamine derivatives do possess some .beta..sub.3 selectivity, side effects of this type have been observed in human volunteers. It is reasonable to expect that these side effects resulted from partial .beta..sub.1 and/or .beta..sub.2 agonism.
More recent developments in this area are disclosed in Ainsworth et al. U.S. Pat. No. 5,153,210, Caulkett et al., U.S. Pat. No. 4,999,377, Alig et al., U.S. Pat. No. 5,017,619, Lecount et al., European Patent 427480 and Bloom et al., European Patent 455006.
Even though these more recent developments purport to describe compounds with greater .beta..sub.3 selectivity over the .beta..sub.1 and .beta..sub.2 activities, this selectivity was determined using rodents, in particular, rats as the test animal. Because even the most highly selective compounds, as determined by these assays, still show signs of side effects due to residual .beta..sub.1 and .beta..sub.2 agonist activity when the compounds are tested in humans, it has become apparent that the rodent is not a good model for predicting human .beta..sub.3 selectivity.
Recently, assays have been developed which more accurately predict the effects that can be expected in humans. These assays utilize cloned human .beta..sub.3 receptors which have been expressed in Chinese hamster ovary cells. See Emorine et al, Science, 1989, 245:1118-1121; and Liggett, Mol. Pharmacol., 1992, 42:634-637. The agonist and antagonist effects of the various compounds on the cultivated cells provide an indication of the antiobesity and antidiabetic effects of the compounds in humans.